Multiple sequence alignment

1. Multiple sequence alignment

2. Multiple sequence alignment: outline [1] Introduction to MSA Exact methods Progressive (ClustalW) Iterative (MUSCLE) Consistency (ProbCons) Structure-based (Expresso) Conclusions: benchmarking studies [3] Hidden Markov models (HMMs), Pfam and CDD [4] MEGA to make a multiple sequence alignment [5] Multiple alignment of genomic DNA

3. Multiple sequence alignment: definition • a collection of three or more protein (or nucleic acid) sequences that are partially or completely aligned • homologous residues are aligned in columns across the length of the sequences • residues are homologous in an evolutionary sense • residues are homologous in a structural sense Page 320

4. ClustalW Note how the region of a conserved histidine (▼) varies depending on which algorithm is used

5. Praline

6. MUSCLE

7. Probcons

8. TCoffee

9. Multiple sequence alignment: properties • not necessarily one “correct” alignment of a protein family • protein sequences evolve... • ...the corresponding three-dimensional structures of proteins also evolve • may be impossible to identify amino acid residues that align properly (structurally) throughout a multiple sequence alignment • for two proteins sharing 30% amino acid identity, about 50% of the individual amino acids are superposable in the two structures Page 320

10. Multiple sequence alignment: features • • some aligned residues, such as cysteines that form • disulfide bridges, may be highly conserved • there may be conserved motifs such as a • transmembrane domain • there may be conserved secondary structure features • there may be regions with consistent patterns of • insertions or deletions (indels) Page 320

11. Multiple sequence alignment: uses • • MSA is more sensitive than pairwise alignment • to detect homologs • BLAST output can take the form of a MSA, • and can reveal conserved residues or motifs • Population data can be analyzed in a MSA (PopSet) • A single query can be searched against • a database of MSAs (e.g. PFAM) • Regulatory regions of genes may have consensus • sequences identifiable by MSA Page 321

12. Multiple sequence alignment: outline [1] Introduction to MSA Exact methods Progressive (ClustalW) Iterative (MUSCLE) Consistency (ProbCons) Structure-based (Expresso) Conclusions: benchmarking studies [3] Hidden Markov models (HMMs), Pfam and CDD [4] MEGA to make a multiple sequence alignment [5] Multiple alignment of genomic DNA [6] Introduction to molecular evolution and phylogeny

13. Multiple sequence alignment: methods Exact methods: dynamic programming Instead of the 2-D dynamic programming matrix in the Needleman-Wunsch technique, think about a 3-D, 4-D or higher order matrix. Exact methods give optimal alignments but are not feasible in time or space for more than ~10 sequences. Still an extremely active field.

14. Multiple sequence alignment: outline [1] Introduction to MSA Exact methods Progressive (ClustalW) Iterative (MUSCLE) Consistency (ProbCons) Structure-based (Expresso) Conclusions: benchmarking studies [3] Hidden Markov models (HMMs), Pfam and CDD [4] MEGA to make a multiple sequence alignment [5] Multiple alignment of genomic DNA [6] Introduction to molecular evolution and phylogeny

15. Multiple sequence alignment: methods Progressive methods: use a guide tree (a little like a phylogenetic tree but NOT a phylogenetic tree) to determine how to combine pairwise alignments one by one to create a multiple alignment. Making multiple alignments using trees was a very popular subject in the ‘80s. Fitch and Yasunobu (1974) may have first proposed the idea, but Hogeweg and Hesper (1984) and many others worked on the topic before Feng and Doolittle (1987)—they made one important contribution that got their names attached to this alignment method. Examples: CLUSTALW, MUSCLE

16. Multiple sequence alignment: methods Example of MSA using ClustalW: two data sets Five distantly related lipocalins (human to E. coli) Five closely related RBPs When you do this, obtain the sequences of interest in the FASTA format! (You can save them in a Word document) Page 321

17. Multidimensional Dynamic Programming • Example: in 3D (three sequences): • 7 neighbors/cells F(i,j,k) = max{ F(i-1,j-1,k-1)+S(xi, xj, xk), F(i-1,j-1,k )+S(xi, xj, - ), F(i-1,j ,k-1)+S(xi, -, xk), F(i-1,j ,k )+S(xi, -, - ), F(i ,j-1,k-1)+S( -, xj, xk), F(i ,j-1,k )+S( -, xj, xk), F(i ,j ,k-1)+S( -, -, xk) }

18. HOW CAN I ALIGN MANYSEQUENCES 2 Globins =>1 Min

19. HOW CAN I ALIGN MANYSEQUENCES 3 Globins =>2 hours

20. HOW CAN I ALIGN MANYSEQUENCES 4 Globins => 10 days

21. HOW CAN I ALIGN MANYSEQUENCES 5 Globins => 3 years

22. HOW CAN I ALIGN MANYSEQUENCES 6 Globins =>300 years

23. HOW CAN I ALIGN MANYSEQUENCES 7 Globins =>30. 000 years Solidified Fossil,Old stuff

24. HOW CAN I ALIGN MANYSEQUENCES 8 Globins =>3 Million years

25. The Choice of an objective function Biological problem that lies in the definition of correctness • Sum of pair, Entropy score, Consistency based, … • The Optimization of that function • Exact Algorithms (Dynamic Programming) • Progressive alignment (ClustalW) • Iterative approaches (SA, GA, …)

26. A A C A A A C A C C A A A C A C A A C Traditional (SP) Tree-Alignment Star-Alignment Input seq Reconstructedseq Missmatches -- A, A, C A 6 1 2 Alignment Costs Traditional A, A, A, C, C A, A, A, C, C A, A, A, C, C

29. The Progressive Multiple Alignment Algorithm (Clustal W)

30. -Greedy Heuristic (No Guarranty). -Fast Making An Alignment Any Exact Method would be TOO SLOW We will use a Heuristic Algorithm. Progressive Alignment Algorithm is the most Popular -ClustalW

31. Progressive Alignment Feng and Dolittle, 1988; Taylor 1989 Clustering

32. Dynamic Programming Using A Substitution Matrix Progressive Alignment

33. Progressive Alignment -Depends on the CHOICE of the sequences. -Depends on the ORDER of the sequences (Tree). • -Depends on the PARAMETERS: • Substitution Matrix. • Penalties (Gop, Gep). • Sequence Weight. • Tree making Algorithm.

34. 1 + 2 1 + 3 1 + 4 2 + 3 2 + 4 3 + 4 Example : Progressive alignment Pairwise Alignment Guide Tree MSA by adding sequences 1 2 3 4 2 3 4 1 1 2 3

35. Progressive alignment (cont.) Sequence Guide Tree 1 2 3 4 5 1 1 2 3 4 5 Distance Matrix: displays distances of all sequence pairs. 2 4 5 3 D = 1 - S UPGMA(unweighted pair group method of arithmetic averages) or Neighbour-Joining method

36. 3 3 3 3 5 5 5 5 1 1 1 1 2 2 2 2 4 4 4 4 UPGMA Clustering (Guide Tree) d d d d ij ij ij ij 1 2 3 4 5 1 0 2 6 9 7 2 0 5 7 7 3 0 5 4 4 0 3 5 0 u 3 v u 0 5 7 3 0 4 v 0 u w u 0 6 w 0 6 0 u 3 4 5 u 0 5 8 7 3 0 5 4 4 0 3 5 0 2 0 . 5 . 5 . 5 . . 8 5 4 0 . . 5 5 3 0

37. Progressive alignment (cont.) Guide Tree Alignment of alignments • Columns - once aligned - are never changed. . . and new gaps are inserted. • Depend strongly on pairwise alignments and the intitial startingsequences • No guarantee that the global optimal solution will be found. • In case of sequences identity less than 25-30%, this approach become much less reliable. 1 2 4 5 2 3 1

39. CLUSTALW (Score=20, Gop=-1, Gep=0, M=1) SeqA GARFIELD THE LAST FA-T CAT SeqB GARFIELD THE FAST CA-T --- SeqC GARFIELD THE VERY FAST CAT SeqD -------- THE ---- FA-T CAT CORRECT (Score=24) SeqA GARFIELD THE LAST FA-T CAT SeqB GARFIELD THE FAST ---- CAT SeqC GARFIELD THE VERY FAST CAT SeqD -------- THE ---- FA-T CAT Progressive Alignment When Doesn’t It Work

40. GARFIELD THE LAST FAT CAT GARFIELD THE LAST FAT CAT GARFIELD THE FAST CAT --- GARFIELD THE FAST CAT GARFIELD THE LAST FA-T CAT GARFIELD THE FAST CA-T --- GARFIELD THE VERY FAST CAT -------- THE ---- FA-T CAT GARFIELD THE VERY FAST CAT GARFIELD THE VERY FAST CAT -------- THE ---- FA-TCAT THE FAT CAT

41. Iterative alignment A B C DE Pairwise distance table Iterate until the MSA doesn’t change (convergence) Guide tree MSA A B C D E

42. The input for ClustalW: a group of sequences (DNA or protein) in the FASTA format

43. Get sequences from Entrez Protein (or HomoloGene)

44. You can display sequences from Entrez Protein in the fasta format

45. When you get a DNA sequence from Entrez Nucleotide, you can click CDS to select only the coding sequence. This is very useful for phylogeny studies.

46. HomoloGene: an NCBI resource to obtain multiple related sequences [1] Enter a query at NCBI such as globin [2] Click on HomoloGene (left side) [3] Choose a HomoloGene family, and view in the fasta format

47. Use ClustalW to do a progressive MSA http://www2.ebi. ac.uk/clustalw/ Fig. 10.1 Page 321

48. Feng-Doolittle MSA occurs in 3 stages [1] Do a set of global pairwise alignments (Needleman and Wunsch’s dynamic programming algorithm) [2] Create a guide tree [3] Progressively align the sequences Page 321

49. Progressive MSA stage 1 of 3: generate global pairwise alignments five distantly related lipocalins best score Fig. 10.2 Page 323

50. Progressive MSA stage 1 of 3: generate global pairwise alignments five closely related lipocalins Start of Pairwise alignments Aligning... Sequences (1:2) Aligned. Score: 84 Sequences (1:3) Aligned. Score: 84 Sequences (1:4) Aligned. Score: 91 Sequences (1:5) Aligned. Score: 92 Sequences (2:3) Aligned. Score: 99 Sequences (2:4) Aligned. Score: 86 Sequences (2:5) Aligned. Score: 85 Sequences (3:4) Aligned. Score: 85 Sequences (3:5) Aligned. Score: 84 Sequences (4:5) Aligned. Score: 96 best score Fig. 10.4 Page 325